Benjamin Studebaker

Bernie Sanders did not have a good night on Tuesday, losing all five contests in Florida, North Carolina, Ohio, Illinois, and even Missouri. The major networks are writing his political obituary–they chose not to broadcast his speech, instead choosing to stand by for more Donald Trump. So how bad is it really?

As some of you might remember, 538 came out with a set of demographic projections to simulate a race between Sanders and Clinton under two different sets of conditions–one in which Sanders trails Clinton by 12 points, and another in which they are tied nationally. I have been checking the results against these projections. Positive numbers indicate a Sanders margin of victory, negative numbers indicate a Clinton margin. Green numbers indicate that Sanders is ahead of the tie projection, red numbers indicate that he is behind. Here’s where we are now:

When we last talked about these figures after Super Tuesday, (when Alabama was in but Kansas wasn’t in yet), I pointed out that Sanders tended to exceed the tie projection in the north, but to massively under-perform in the south. Large Clinton margins in Florida and North Carolina are consistent with that. The projections expected Sanders to lose Illinois even in a tied contest, and he lost it by a smaller margin than the tie projection indicated. Missouri was expected to be close, and while Sanders is surely disappointed to lose it narrowly rather than win it narrowly, this does not make much difference to the delegate count. But what about Ohio?

The tie projection called for Sanders to lose Ohio by single digits, and he lost it by double digits. It’s nearly in line with the 12 point gap projection, and those numbers are not consistent with winning. Outside the south, there is only one other state where Sanders performed like he was down by 12–Massachusetts. Each time he loses a northern state by figures in line with the down 12 projections, his margin of error gets markedly smaller.

But Bernie Sanders still thinks he can win. He says:

With more than half the delegates yet to be chosen and a calendar that favors us in the weeks and months to come we remain confident that our campaign is on a path to win the nomination.

How can Bernie Sanders think this? He’s noticed something about Ohio that explains why it wasn’t a close contest like Illinois or Michigan even though demographically it’s a similar kind of state–the turnout figures were completely different.

In Michigan, the democrats and the republicans both turned out nearly the same number of voters:

Sanders won Michigan, but by less than the demographic tie projection.

In Illinois, the democrats did much better than the republicans on turnout:

Sanders didn’t win Illinois, but he exceeded the demographic tie projection, which is from his point of view more important.

But in Ohio? The democrats got slammed on turnout:

Ohio is an open primary, which means that you can choose which party’s primary you prefer to participate in. Sanders is likely thinking that many people chose to vote for home state governor John Kasich in an attempt to stump the Trump. While Kasich did win Ohio, this effort is clearly failing nationwide–last night Trump won every state except Ohio, forcing Marco Rubio to drop out. Rejuvenated by his success in Ohio, Kasich now plans to stay in, and this will continue to prevent the republican establishment’s Trump ceiling theory from being tested (this theory claims that Trump has a “ceiling” of about 40% nationally, and that if the republicans field a single anti-Trump candidate, that candidate can defeat Trump). So the Kasich voters have not really accomplished anything–indeed, by keeping Kasich in the race, they have arguably aided and abetted Trump. In the meantime, they have made it harder for Sanders to look viable in future contests. Ohioans are just facilitating a Clinton/Trump race, which for most Americans is a lesser of two evils choice between the most unpopular democratic candidate in history (-12.0 favorability net score) and the most unpopular republican candidate in history (-29.1 favorability).

But if you’re Bernie Sanders, you don’t fall into the pit of despair. You look at Ohio and you think “Ohio is the exception, Illinois and Michigan are the rule–John Kasich isn’t the governor of the rest of the states on the calendar”. You look ahead at that calendar, and you see that even if you’re down 12 points nationally, you stand to win 7 of the next 8 states, and if you hit or exceed the tie projection going forward, you could win as many as 13 in a row:

There are no states that are unambiguously against Sanders until Maryland on April 26. Sanders needs big margins of victory in the western states followed by wins in New York, Pennsylvania, and California. Can he do it? It’s far from probable, but it might be possible.

For the record, I don’t think Bernie Sanders is going to beat Hillary Clinton. I don’t think enough democrats recognize how much this primary matters or how electable Sanders is. But there are so many people who are happy to tell you that Sanders is finished, that he has no chance at all. I don’t think he has no chance at all. At this point his chance is quite small, but when was he ever the favorite to win against Clinton? Sanders has always been a dark horse. The horse may be a shade darker now, but it still has legs and it continues to run. There’s certainly no cost to betting on a dark horse in the primaries.